# Articles tagged with: research

> Anthus builds self-aligning AI systems: custom models, agent harnesses, evaluation loops, and unattended automation with a human in the loop, grounded in production operations since 2007.

- [Can Averaging Reduce Noise in an AI Decision Model?](https://anth.us/articles/can-averaging-reduce-noise/index.md): We asked Jev the same question ten times in one request and expected ten identical answers. They split, four votes to six. Kahneman, Sibony and Sunstein call that kind of variation noise. Averaging the copies didn't make Jev more accurate, but it cut changed answers on a rerun by about a third. (2026-10-01)
- [Decision Models Are Not Calculators](https://anth.us/articles/decision-models-are-not-calculators/index.md): Jev chose one in every ordering of a fair-die question; Kev leaned toward one, and Laya overwhelmingly chose six. Benford's Law remains a hypothesis. (2026-09-25)
- [The One-Word Test: How Jev and Laya Read Gender, Race and Age](https://anth.us/articles/one-word-test/index.md): Two thousand real bios per job, first names blanked, one pronoun swapped, both decision models asked twice. Laya's verdict moved on 8 to 18 bios in 100, Jev's on 1 to 4, almost always toward the stereotype. Here is the method, the jobs, the shortlist, and what names and ages do. (2026-09-22)
- [We Told the AI She Was a Woman. It Demoted Her](https://anth.us/articles/encoding-prejudice/index.md): A new kind of AI, the decision model, makes instant decisions about people: who gets an interview, whose claim gets flagged. It's cheap and fast, and companies are adopting it without checking what it does. We checked. It quietly ranks women lawyers below men, and every company using it makes the s… (2026-09-22)
- [Can You Fix It? Gating, Averaging and Fine-Tuning Against a Gendered Verdict](https://anth.us/articles/can-you-fix-it/index.md): A decision model's verdict moves on "she". We put a gate on the learning loop that rejects any question whose answer moves with the pronoun. It kept out the obvious ones and still made a hiring shortlist worse, because a question can ignore the pronoun and read gender from the words. Here's what wo… (2026-09-22)
- [Jev vs Laya: Same Labels, Same Questions, One Variable](https://anth.us/articles/jev-vs-laya/index.md): We ran two decision models, Jev and the open-weights Laya, through one harness where only the engine changes. Jev led by 4.7 points alone and 6.8 with our feedback layer, and then fine-tuning Laya on the same 140 labels beat both. (2026-09-21)
- [Fine-Tuning Jev: You Can't. Here's What Gets You the Same Effect](https://anth.us/articles/fine-tuning-jev/index.md): You can't fine-tune Jev, or any hosted decision model: TypeSafe serves the same weights to everyone. So we kept it frozen and adapted the questions and a small fitted head around it. On 140 labels, one new plain-English question bought 10 points of accuracy, and it finds the hidden pattern about a… (2026-09-21)
- [Distilling an Aligned Jev System into a Classifier You Own](https://anth.us/articles/distilling-jev-into-a-classifier/index.md): A Jev decision model aligned with 140 human labels taught a 66M-parameter DistilBERT that scored 0.912 against the human label, two points above its teacher, at 5.6 to 15 ms an item. The accuracy is the least general part; the per-slice ship gate is the part to copy. (2026-09-21)
- [Can You Trust Jev's Confidence?](https://anth.us/articles/can-you-trust-jev-confidence/index.md): A decision model's confidence should mean something. Jev's does: higher really does mean more likely right. But on 8,801 labeled examples they ran overconfident, and how you phrase the request changes the pattern. Calibration fixes most of it. (2026-09-19)
- [Decision Models: Making Decisions Instead of Generating Text](https://anth.us/articles/making-decisions-instead-of-generating-text/index.md): Jev is a decision model: it turns a scorecard’s bounded questions into the model’s native output, making high-volume QA faster, cheaper, and easier to operate. (2026-09-17)
- [The Turn Detection Trap: When 100% Accuracy is Wrong](https://anth.us/articles/domain-specific-turn-detection/index.md): We fine-tuned MobileBERT for turn detection and achieved 100% accuracy. Then we realized we hadn't solved turn detection at all—we had just built a very expensive punctuation detector. (2025-11-22)
- [The Dominance of Ones: A Handy Quirk of Numbers](https://anth.us/articles/the-dominance-of-ones/index.md): From money launderers to breeding rabbits, this fascinating quirk of numbers is lurking on the sidelines, ready to surprise us. (2023-11-26)